Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/paulnsorensen/easy-cheese/python-authoringnpx skills add paulnsorensen/easy-cheese --skill python-authoringgit clone --depth 1 https://github.com/paulnsorensen/easy-cheeseWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00088 | $0.02015 |
| Opus 5 | $0.00044 | $0.01007 |
| Sonnet 5 | $0.00018 | $0.00403 |
| Haiku 4.5 | $0.00009 | $0.00201 |
Grade A, and why
python-authoring scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Authoring Python
Produce the smallest readable Python change that satisfies the request and matches easy-cheese.
This is a repository-local skill. Keep it under .agents/skills/python-authoring/; do not mirror it into the published skills/ tree or add agents/openai.yaml. Renaming this directory requires updating the matching !.agents/skills/<dir>/ line in .gitignore — that path list is an allowlist, so a renamed dir without its own line is silently untracked.
Work in this order
- Read the root configuration, owning skill, target exports, immediate callers, and nearby conventions.
- Keep every changed line traceable to the request. Do not invent orchestration, compatibility layers, abstractions, dependencies, or future flexibility.
- Put code in the owning package and preserve the bundle boundary.
- Validate untrusted input once at the boundary, then work with typed trusted data.
- Choose the clearest succinct Python construct; do not compress code until it becomes harder to read.
- Remove only slop introduced by the change and code that the change orphaned.
- Type-check changed files with basedpyright, run targeted tests, rebuild affected bundles, then run
just check.
Keep runtime code stdlib-first
- Target Python 3.12. Use its language and typing features directly; do not add compatibility code for older versions.
- Treat the standard library as the default dependency budget for bundled helpers under
src/. Any third-party runtime dependency must be pure Python and admitted torequirements/runtime.txt; bundle builds resolve internal and external hashes into an ephemeral requirements file. - Reuse the existing JSON-first, optional-YAML manifest boundary instead of importing PyYAML into new bundled modules.
- Treat configured third-party imports as surface-specific exceptions: PyYAML for existing validators, docs tooling, and manifest/test paths; pytest for tests. A new dependency requires an explicit package and CI decision.
- Prefer
argparse,json,pathlib,tempfile,shutil,zipfile,collections,itertools, andcontextlibover hand-written equivalents.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 102 lines · 88 tokens per session scan A 792f5eaa315c
python-authoring is a skill published in the GitHub repository paulnsorensen/easy-cheese (18 stars, last pushed 3d ago), licensed MIT. It adds 88 tokens to every session and 2,015 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
agentic-development
Build AI agents with Pydantic AI (Python) and Claude SDK (Node.js).
django-idioms
Django ORM, class-based views, DRF serializers, migrations, pytest-django. For Python see python-idioms.
python-expert
Use for Python code changes, FastAPI/Django/Flask services, packaging, typing, async, pytest, data scripts, dependency hygiene, or Python performance fixes.
unreal-pcg-python
Guide for Unreal Engine 5.x PCG (Procedural Content Generation) Python integration. Covers the PCGPythonInterop plugin, the Execute Python Script node, PCG Python API (PCGComponent, PCGBlueprintElement, PCGSpatialData, PCGPointData), editor automation, custom PCG nodes via Python, and known limitations. Use when the…
upgrade-claude-agent-sdk
Upgrade the claude-agent-sdk dependency, review SDK changes for integration impact, expose new features to users via Settings or New Task dialog, update documentation, and ship on a feature branch. This skill should be used when the user mentions upgrading, updating, or bumping the claude-agent-sdk, or when reviewing…
first-plan-lens-python
Stack lens para Python. Use durante Discovery quando pyproject.toml, setup.py ou requirements.txt for detectado. Cobre FastAPI, Django, Flask, Litestar, Celery workers, data apps, packaging src/ vs flat.